Blue Origin's BE-7 engine

[hkultala]

It would really make sense to use this for a new 3rd stage on top of New Glenn 9x4, for high energy launches:

4 BE-3Us in the second stage give almost 400 tonnes of thrust, which gives very small gravity losses with huge second stage and allows lifting 70 tonnes to LEO, but the tanks of this huge second stage are quite heavy and the tank weight eats payload to higher orbits.

But, lets add a smallish third stage with single BE-7.

This stage would only stage at orbital speed, eliminating gravity loss, so the very weak engine would not matter much (only losing small amount of performance due to less oeberth effect)

70 tonnes of initial mass , 30 tonnes of propellant, 40 tonnes of final mass (something like 3.5 tonnes of mass for the stage  and 36.5 tonnes of payload) would get from earth to GTO.

Or for TLI, 28-tonne payload would give delta-v of 3 km/s for this stage. As the staging would happen slightly higher at slightly elliptic orbit due to only 61.5 tonnes of payload weight lifted by the second stage, this should be enough for TLI.

This is about the same than what SLS Block 1 can lift to TLI.

Or, towards Mars: 22-tonne payload would mean delta-v of 3.5 km/s for this stage.
the remaining 300 m/s is easily done be earlier stages, due to only 55.5 tonnes (instead of 70 tonnes) of weight lifted by the second stage.

This is much more than Falcon Heavys 17 tonnes towards Mars.

Could even launch decent-size probes towards outer solar system without slow complicated gravity slings, for example 5-tonne payload would get delta-v of 6.8 km/s.

and as the second stage would only need lift 38.5 tonnes, the staging would happen at considerably higher that LEO, total delta-v might be over LEO+8 km/s

And this is for mass about 6 times bigger than the Voyager probes.


The stage would have quite a long burn time (about 45 minutes)

This stage should also be quite cheap.

[sdsds]

]With rocket engines, boring is good. To that end, please enjoy this 1,030-second (17+ min!) BE-7 engine burn. This test represents the Apogee Raise Maneuver or ARM burn for our Blue Moon Mark 1 Lunar lander

A 17 minute burn can't really be approximated as a single impulse maneuver. Do they lose much to something like cosine losses, or by always thrusting along the velocity vector passing through perigee do those effectively cancel out?

[hkultala]

Does the BE-7 less expensive than the RL-10?  If so can it replace the RL-10 to cut costs for ULA who uses the RL-10?

Due to much lower thrust, 5 BE-7's would be needed to replace the two RL-10's with BE-7.

And those 5 BE-7's might physically not fit under the second stage of Vulcan.

And those 5 BE-7's would probably not be very cheap either even though single BE-7 is probably considerably cheaper than single RL-10.

It is not designed to be a second stage engine of a booster rocket, it is designed to be practically only used as space tug engine / engine for space-only stage.

[envy887]

Does the BE-7 less expensive than the RL-10?  If so can it replace the RL-10 to cut costs for ULA who uses the RL-10?

Due to much lower thrust, 5 BE-7's would be needed to replace the two RL-10's with BE-7.

And those 5 BE-7's might physically not fit under the second stage of Vulcan.

And those 5 BE-7's would probably not be very cheap either even though single BE-7 is probably considerably cheaper than single RL-10.

It is not designed to be a second stage engine of a booster rocket, it is designed to be practically only used as space tug engine / engine for space-only stage.

Centaur would almost certainly benefit from having even more thrust, which would tend to push ULA towards BE-3U rather than BE-7, if they wanted to switch engines.

[Brigantine]

Increasin the ratio of oxygen to hydrogen increase the thrust in some rocket engins.
Increasin the ratop of hydrogen to oxygen increas the specific impulse.

Trying to process this... So is it fair to say the mass rate of LH₂ hits the limit, but you can still pump LOx faster for more thrust (but with diminishing returns)

So when you're fighting gravity drag or there's an Oberth advantage to a shorter burn time, you hit that switch.

And further, that 44 kN thrust and 4.5 km/s ISP are both maxima which do not occur at the same time as each other. Now it sounds like even more advantage to having several BE-7's on e.g. GS3, if the engine itself is not heavy.

[hkultala]

Does the BE-7 less expensive than the RL-10?  If so can it replace the RL-10 to cut costs for ULA who uses the RL-10?

Due to much lower thrust, 5 BE-7's would be needed to replace the two RL-10's with BE-7.

And those 5 BE-7's might physically not fit under the second stage of Vulcan.

And those 5 BE-7's would probably not be very cheap either even though single BE-7 is probably considerably cheaper than single RL-10.

It is not designed to be a second stage engine of a booster rocket, it is designed to be practically only used as space tug engine / engine for space-only stage.

Centaur would almost certainly benefit from having even more thrust, which would tend to push ULA towards BE-3U rather than BE-7, if they wanted to switch engines.

Yep, especially for maximizing payload to LEO.

For high-energy orbits, the benefits would be smaller as staging happens later and there is slightly less mass to lift, so less gravity losses anyway.

[hkultala]

Increasin the ratio of oxygen to hydrogen increase the thrust in some rocket engins.
Increasin the ratop of hydrogen to oxygen increas the specific impulse.

Trying to process this... So is it fair to say the mass rate of LH₂ hits the limit, but you can still pump LOx faster for more thrust (but with diminishing returns)

So when you're fighting gravity drag or there's an Oberth advantage to a shorter burn time, you hit that switch.

The difference of benefits of Oeberth effect are practically never as big that you'd want to sacrifice isp for thrust.

You only start having considerable losses for not utilizing oeberth effect if you have really lousy T/W and in that case, something like 10% more thrust is not going to save anything (you lose much more with the decrease in ipc)

But for fighting gravity, immediately after liftoff or staging when your fuel tank is full of propellant, you need all the thrust you can get to minimize gravity losses and something like 10% thrust boost (for running more oxygen rich mixture) for the first minute can be a big help in reducing the gravity losses.

And then when the gravity losses get smaller start gradually decreasing the flow of oxygen to get better isp (and less thrust).

BE-7 with the dual expander cycle might allow easier variable mixture ratio operation than other rocket engines, though as BE-7 is only meant to be used in orbit, the thrust does not matter much and it will probably always be running at mixture ratio optimized for isp, not thrust.

[sdsds]

Getting the optimal number of low-thrust engines (BE-7s in this case) underneath a stage seems tricky. If they had 4x the number of BE-7 engines they could both cut the burn time from 17 min down to ~4, and potentially complete the burn even with an engine out. (Assuming propellant loading for that case.) As it is, since they're calling this an Apogee Raise Maneuver, are they implicitly dividing the TLI requirement between two burns. Is there refilling of propellant between the two?

[Brigantine]

There was a render (of dubious origin, but seemingly plausible) of a GS3 with I calculated 40 tons of prop and a single BE-7. That's over an hour of burn time. Generally Oberth doesn't matter much because mission planners simply don't push it that far?

It's possible that "GS3" is more of a single-use space tug than a rocket stage, that can last weeks in space and split the burn over many orbits and then help with lunar capture. I guess even if it could only last 24 or 48 hours you could still do several apogee raise maneuvers before TLI.

Do you think this is real?
<image>

[deltaV]

It would really make sense to use this for a new 3rd stage on top of New Glenn 9x4, for high energy launches:

...

But, lets add a smallish third stage with single BE-7.

This stage would only stage at orbital speed, eliminating gravity loss, so the very weak engine would not matter much (only losing small amount of performance due to less oeberth effect)

...

The stage would have quite a long burn time (about 45 minutes)

I agree that 9x4 needs a third stage to do well for high energy orbits and that BE 7 is a good engine for this. I disagree on a detail, namely how many BE-7s to use. Your 45 minute burn time is about half a low earth orbit so the burns would be very far from impulsive and the Oberth effect losses would presumably be substantial. I suspect 2-4 BE-7s, which would give similar thrust to weight ratio as the Vulcan, Atlas V, and Delta IV upper stages, would likely trade better.

[hkultala]

It would really make sense to use this for a new 3rd stage on top of New Glenn 9x4, for high energy launches:

...

But, lets add a smallish third stage with single BE-7.

This stage would only stage at orbital speed, eliminating gravity loss, so the very weak engine would not matter much (only losing small amount of performance due to less oeberth effect)

...

The stage would have quite a long burn time (about 45 minutes)

I agree that 9x4 needs a third stage to do well for high energy orbits and that BE 7 is a good engine for this. I disagree on a detail, namely how many BE-7s to use. Your 45 minute burn time is about half a low earth orbit so the burns would be very far from impulsive and the Oberth effect losses would presumably be substantial. I suspect 2-4 BE-7s, which would give similar thrust to weight ratio as the Vulcan, Atlas V, and Delta IV upper stages, would likely trade better.

There would be some losses form burning at higher altitude(lower speed), but they would not be so great.

GTO has orbital period of about 10.5 hours, so 5.25 hours of ascent time.

This 45 minutes burn time is about 15% of that ascent time. Most of the burn would still be happening at very low altitude, high speed.

Se the losses would not be great.

Also, engines are the most expensive parts of rocket stages, so even though adding another engine might improve performance slightly, cost-wise it might not make sense.

And instead of adding engines, we could also make the stage even lighter, staging at speed higher than orbital speed even for GTO missions.

And Centaur and DCSS stage at much lower so they are really fighting against gravity, comparison to them is not a good comparison.

This third stage would only be staging at about orbital velocity.

[Brigantine]

not be so great.

GTO has orbital period of about 10.5 hours, so 5.25 hours of ascent time.

This 45 minutes burn time is about 15% of that ascent time. Most of the burn would still be happening at very low altitude, high speed.

Se the losses would not be great.

IMO the relevant back-of-the-napkin metric is how many hours it takes until you double your radius. From e.g. 6371+200 km to 6371+6771 km. In potential energy terms, that's half way to infinity.

What's the cost of a BE-7 vs four BE-3U's?

The single-BE-7 approach makes sense if you're biting the bullet on stage life extension, doing apogee-raise maneuvers before the main burn.

[ZachF]

Does the BE-7 less expensive than the RL-10?  If so can it replace the RL-10 to cut costs for ULA who uses the RL-10?

Due to much lower thrust, 5 BE-7's would be needed to replace the two RL-10's with BE-7.

And those 5 BE-7's might physically not fit under the second stage of Vulcan.

And those 5 BE-7's would probably not be very cheap either even though single BE-7 is probably considerably cheaper than single RL-10.

It is not designed to be a second stage engine of a booster rocket, it is designed to be practically only used as space tug engine / engine for space-only stage.

I want Blue to make a high pressure FFSC vacuum hydrolox engine for space tugs/3rd stage with a >500 expansion ratio. Something with truly stupid-high ISP. c'mon Blue... do it. Then give it a spherical balloon tank to minimize empty mass. Go all out

[hkultala]

Does the BE-7 less expensive than the RL-10?  If so can it replace the RL-10 to cut costs for ULA who uses the RL-10?

Due to much lower thrust, 5 BE-7's would be needed to replace the two RL-10's with BE-7.

And those 5 BE-7's might physically not fit under the second stage of Vulcan.

And those 5 BE-7's would probably not be very cheap either even though single BE-7 is probably considerably cheaper than single RL-10.

It is not designed to be a second stage engine of a booster rocket, it is designed to be practically only used as space tug engine / engine for space-only stage.

I want Blue to make a high pressure FFSC vacuum hydrolox engine for space tugs/3rd stage with a >500 expansion ratio. Something with truly stupid-high ISP. c'mon Blue... do it. Then give it a spherical balloon tank to minimize empty mass. Go all out

Makes no sense when they already have dual-expander engine that gives practically all the benefits of FFSC in small upper stage engine, but is simpler, cheaper and more reliable than FFSC.

Dual expander cycle engine like BE-7 is already a full-flow closed cycle engine. That gives it all the same benefits that FFSC would have. FFSC would not give any better isp than dual expander.

FFSC makes no sense in space tug engine that is small enough for closed expander cycle to work.

BE-7 already has higher specific impulse than any current operational rocket engine, though new RL10E-1 may surpass it by about 1s quite soon. (neither RL10B-X or RD-0146D are operational, RL10B-X was cancelled and RD-0146D is vaporware, and RL10B-2 and RL10C-2-1 were retired)

And BE-7 could be tuned to have higher isp by
1) increasing the bell size
2) using more hydorgen-rich mixture ratio

but these have their negatives, bigger engine bell is extra weight and size and richer mixture means worse impulse density (more tank volume => tank weight) and also lower thrust.

The engineers of Blue Origin are not stupid, BE-7 has very high ipc (460s) instead of super-high isp (>470s) because they have targeted the best overall comphromize between tank size, weight, trust and ipc.

[deltaV]

Due to much lower thrust, 5 BE-7's would be needed to replace the two RL-10's with BE-7.

And those 5 BE-7's might physically not fit under the second stage of Vulcan.

I can't find a good source for BE-7's nozzle diameter but https://forum.nasaspaceflight.com/index.php?topic=61201.msg2740796#msg2740796 suggests 0.94 m and pictures of BE-7 seem consistent with that. Centaur V is 5.4 m diameter. According to https://www.engineeringtoolbox.com/smaller-circles-in-larger-circle-d_1849.html with those diameters 23 engines would fit. Of course gimbaling would reduce that number a bit, but there's room for a lot more than 5. (Engine height is presumably easier with BE-7s than RL-10s since BE-7s are smaller.)

BTW I'm not saying that a Centaur V re-engined with BE-7s is likely. My point is only that BE-7s being too big is not why such a stage is unlikely.

[ZachF]

Does the BE-7 less expensive than the RL-10?  If so can it replace the RL-10 to cut costs for ULA who uses the RL-10?

Due to much lower thrust, 5 BE-7's would be needed to replace the two RL-10's with BE-7.

And those 5 BE-7's might physically not fit under the second stage of Vulcan.

And those 5 BE-7's would probably not be very cheap either even though single BE-7 is probably considerably cheaper than single RL-10.

It is not designed to be a second stage engine of a booster rocket, it is designed to be practically only used as space tug engine / engine for space-only stage.

I want Blue to make a high pressure FFSC vacuum hydrolox engine for space tugs/3rd stage with a >500 expansion ratio. Something with truly stupid-high ISP. c'mon Blue... do it. Then give it a spherical balloon tank to minimize empty mass. Go all out

Makes no sense when they already have dual-expander engine that gives practically all the benefits of FFSC in small upper stage engine, but is simpler, cheaper and more reliable than FFSC.

Dual expander cycle engine like BE-7 is already a full-flow closed cycle engine. That gives it all the same benefits that FFSC would have. FFSC would not give any better isp than dual expander.

FFSC makes no sense in space tug engine that is small enough for closed expander cycle to work.

BE-7 already has higher specific impulse than any current operational rocket engine, though new RL10E-1 may surpass it by about 1s quite soon. (neither RL10B-X or RD-0146D are operational, RL10B-X was cancelled and RD-0146D is vaporware, and RL10B-2 and RL10C-2-1 were retired)

And BE-7 could be tuned to have higher isp by
1) increasing the bell size
2) using more hydorgen-rich mixture ratio

but these have their negatives, bigger engine bell is extra weight and size and richer mixture means worse impulse density (more tank volume => tank weight) and also lower thrust.

The engineers of Blue Origin are not stupid, BE-7 has very high ipc (460s) instead of super-high isp (>470s) because they have targeted the best overall comphromize between tank size, weight, trust and ipc.

You use FFSC so you can get ultra high expansion ratios without gigantic engine bells. Double chamber pressure with same sized bell = double expansion ratio.

Dual closed expander still has almost the same power limitations as closed hydrogen fuel expander, as like 80% of the heat capacity in hydrolox is on the fuel side.

Also, i never said me wanting to see someone build the ultimate chemical vacuum engine was purely rational, i just want to see someone do it from an engineering perspective.



 

[hkultala]

Does the BE-7 less expensive than the RL-10?  If so can it replace the RL-10 to cut costs for ULA who uses the RL-10?

Due to much lower thrust, 5 BE-7's would be needed to replace the two RL-10's with BE-7.

And those 5 BE-7's might physically not fit under the second stage of Vulcan.

And those 5 BE-7's would probably not be very cheap either even though single BE-7 is probably considerably cheaper than single RL-10.

It is not designed to be a second stage engine of a booster rocket, it is designed to be practically only used as space tug engine / engine for space-only stage.

I want Blue to make a high pressure FFSC vacuum hydrolox engine for space tugs/3rd stage with a >500 expansion ratio. Something with truly stupid-high ISP. c'mon Blue... do it. Then give it a spherical balloon tank to minimize empty mass. Go all out

Makes no sense when they already have dual-expander engine that gives practically all the benefits of FFSC in small upper stage engine, but is simpler, cheaper and more reliable than FFSC.

Dual expander cycle engine like BE-7 is already a full-flow closed cycle engine. That gives it all the same benefits that FFSC would have. FFSC would not give any better isp than dual expander.

FFSC makes no sense in space tug engine that is small enough for closed expander cycle to work.

BE-7 already has higher specific impulse than any current operational rocket engine, though new RL10E-1 may surpass it by about 1s quite soon. (neither RL10B-X or RD-0146D are operational, RL10B-X was cancelled and RD-0146D is vaporware, and RL10B-2 and RL10C-2-1 were retired)

And BE-7 could be tuned to have higher isp by
1) increasing the bell size
2) using more hydorgen-rich mixture ratio

but these have their negatives, bigger engine bell is extra weight and size and richer mixture means worse impulse density (more tank volume => tank weight) and also lower thrust.

The engineers of Blue Origin are not stupid, BE-7 has very high ipc (460s) instead of super-high isp (>470s) because they have targeted the best overall comphromize between tank size, weight, trust and ipc.

You use FFSC so you can get ultra high expansion ratios without gigantic engine bells. Double chamber pressure with same sized bell = double expansion ratio.

Just making the bell bigger would typically still be the better solution. And in expansion cycle engine it would also give more power for the pumps => more pressure also, though dual expansion cycle makes this more complicated.

And the nozzle of BE-7 as it is now is quite small, considerable increase in expansion ratio could be possible without making it "gigantic".

The first version of BE-7 probably has that quite small nozzle because it's made for the blue moon lander and there bigger bell would also mean longer legs.

If they later make a space tug/upper stage-optimized version of BE-7, that might have a bigger nozzle.

Dual closed expander still has almost the same power limitations as closed hydrogen fuel expander, as like 80% of the heat capacity in hydrolox is on the fuel side.

.. but BE-7 is much smaller engine than biggest closed cycle expander engines. It is nowhere near the limits where expander cycle engines stop working, which practically mean that it can run at higher pressures than bigger expander cycle engines such as Vinci.